Michalewicz, Z., "Genetic algorithms, numerical optimization, and constrains,". In. L. Eshelman, ed., Proc. of the 6 th Int. Conf. Genetic Algorithms, pp.151158, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....computation techniques have received a considerable attention regarding their robustness in solving complex optimization problems involving nondifferentiable and discontinuous nonlinearity and high dimension. There have been attempts to address this problem, for example the works in [1 3][5 6]. These methods can be divided into two groups penalty function based approaches and heuristics based approaches. However, the boundary between these two groups is not clear. Some penalty function methods were developed using heuristic rules to penalize unfeasible solutions. Heuristics based ....

Michalewicz, Z., "Genetic algorithms, numerical optimization, and constrains,". In. L. Eshelman, ed., Proc. of the 6 th Int. Conf. Genetic Algorithms, pp.151158, 1995.

....to save computation time, the mutation is retried only ten times and then ignored, leaving the object variable in its original state within the parameter bounds. B. Experimental Results and Discussions Thirteen benchmark functions were used. The first 12 were taken from [14] and the 13th from [15]. The details, including the original sources, of these functions are listed in the Appendix. Problems g02, g03, g08, and g12 are maximization problems. They were transformed into minimization problems using . For each of the benchmark problems, 30 independent runs were performed using a (30, ....

....are based on feasible solutions only. All equality constraints have been converted into inequality constraints, using the degree of violation . As a result of this approximation, some results might be better than the optimum. However, the tolerated violation is more stringent than others [15] where was used. In comparison with the latest results in the literature [14] the results in Table II are significantly better for all but one problem. While 70 20 000 function evaluations were used for each problem and only 20 runs were carried out in [14] for Experiment 2 in [14] which gave ....

[Article contains additional citation context not shown here]

Z. Michalewicz, "Genetic algorithms, numerical optimization and constraints, " in Proc. 6th Int. Conf. Genetic Algorithms, L. J. Eshelman, Ed. San Mateo, CA: Morgan Kaufman, July 1995, pp. 151--158.

....an EA s capability to handle various problem domain characteristics. These suites incorporate relevant search space features which should be addressed by a particular EA instantiation. For example, De Jong [2] suggests five single objective optimization test functions (F1 F5) and Michalewicz [13] suggests five single objective constrained optimization test functions(G1 G5) Whitley et al. 21] and Goldberg et al. 8] offer other test suite functions. Whitley et al. also offer general test suite guidelines which include incorporating real world problems, problems ranging in difficulty ....

....Jong s test bed includes functions with the following characteristics [7] continuous and discontinuous, convex and nonconvex, unimodal and multimodal, quadratic and nonquadratic, low and high dimensionality, and deterministic and stochastic. Michalewicz s test bed addresses the following issues [13]: type of objective function, number of decision variables and constraints, types of constraints, number of active constraints at the function s optimum, and the ratio between the feasible and complete search space size. Particular EA instantiations are then subjected to test beds like these and ....

Michalewicz, Zbigniew. "Genetic Algorithms, Numerical Optimization, and Constraints." Proceedings of the Sixth International Conference on Genetic Algorithms, edited by Larry J. Eshelman. 151--158. San Mateo CA: Morgan Kaufmann Publishers, Inc., 1995.

....if both parents are feasible, especially in highly constrained problems where the constraint is likely to be active. There has been a body of work published in evolutionary computation on handling constraints (the most recent comprehensive treatment is found in [29] In particular, Michalewicz [30 33] and Smith [34 35] have worked on using penalty functions to effectively and efficiently guide evolutionary search to feasible, optimal (or near optimal) final solutions. The penalty function below uses the notion of distance of the solution from feasibility (the R(x) R o term) and a nonlinear ....

Z. Michalewicz, "Genetic Algorithms Numerical Optimization and Constraints", Proceedings of the 6th International Conference on Genetic Algorithms, 1995, pp 151-158.

....crossover should alter the clock periods of the offsprings. 5.4.2 Constraint Handling The occurrence of LTCs in the ETPN makes constraint handling an important part of the algorithm. A number of methods have been proposed for constraint handling, specially for numerical optimization problems [Mic 95] In this work a new approach is taken. A penalty function that changes the size of the penalty, considering the number of infeasible chromosomes in the population, is introduced. The purpose of the penalty function is to keep the number of infeasible chromosomes in the population constant. The ....

Z. Michalewicz, "Genetic Algorithms, Numerical Optimization, and Constraints," in Proc. 6th International Conference on Genetic Algorithms, Pittsburgh, July 15-19, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC